Oahu LiDAR QA Report May 16, 2014 Page 1 of 30
LiDAR Quality Assurance (QA) Report QA Report for Oahu LiDAR
Produced for National Oceanic and Atmospheric Administration NOAA Contract: W912P9-10-D-0534
Task Order: 0016
Report Date: May 16, 2014
SUBMITTED BY:
Dewberry
1000 North Ashley Drive Suite 801 Tampa, FL 33602 813.225.1325 SUBMITTED TO:
NOAA 2234 S Hobson Ave Charleston, SC 29405 843.740.1323
Oahu LiDAR QA Report May 16, 2014 Page 2 of 30
Table of Contents
Executive Summary .............................................................................................................................. 4
Deliverables Summary for Oahu LiDAR ......................................................................................... 6
Overview ................................................................................................................................................ 7
LiDAR Analysis ..................................................................................................................................... 7
LiDAR Quantitative Review ............................................................................................................. 7
Survey Vertical Accuracy Checkpoints ......................................................................................... 7
Vertical Accuracy Results – Swath LiDAR .................................................................................. 11
Vertical Accuracy Results – Classified LiDAR ........................................................................... 12
LiDAR Completeness Review ......................................................................................................... 13
Point Count/Elevation Analysis ................................................................................................. 14
LiDAR Qualitative Review .............................................................................................................. 15
Aggressive Misclassification ....................................................................................................... 15
Artifacts ........................................................................................................................................ 16
Corn Rows .................................................................................................................................... 18
Vegetation Artifacts in Ground ................................................................................................... 19
LiDAR Recommendation............................................................................................................... 20
Breakline Analysis ............................................................................................................................. 20
Breakline Data Overview ............................................................................................................... 20
Breakline Qualitative Review ........................................................................................................ 20
Breakline Quantitative Review ...................................................................................................... 20
Topology .......................................................................................................................................... 21
Monotonicity ................................................................................................................................ 21
Connectivity ................................................................................................................................. 21
Breakline Recommendation ........................................................................................................... 21
Hydro-flattened Digital Elevation Model Analysis .......................................................................... 21
Overview .......................................................................................................................................... 21
DEM Quantitative Review .............................................................................................................. 22
Vertical Accuracy Results – Digital Elevation Model (DEM) ................................................... 22
Qualitative Review .......................................................................................................................... 23
Artifacts ........................................................................................................................................ 23
DEM Recommendation .................................................................................................................. 23
Metadata.............................................................................................................................................. 23
Metadata Recommendation ........................................................................................................... 23
Other Comments................................................................................................................................. 24
Recommendations Summary............................................................................................................. 24
Oahu LiDAR QA Report May 16, 2014 Page 3 of 30
LiDAR: ............................................................................................................................................. 24
Breaklines: ....................................................................................................................................... 24
DEMs: .............................................................................................................................................. 24
Metadata: ......................................................................................................................................... 24
APPENDIX A ─ ACCEPTANCE CRITERIA ...................................................................................... 25
LiDAR Data Acquisition Acceptance Criteria ............................................................................ 25
LiDAR Accuracy Acceptance Criteria ......................................................................................... 25
Geographic Coverage and Continuity Acceptance Criteria ....................................................... 25
Spatial Reference Framework ..................................................................................................... 25
Deliverables .................................................................................................................................. 25
Usability Acceptance Criteria ......................................................................................................26
Breakline Acceptance Criteria.....................................................................................................26
Appendix B- QA/QC Checklists ........................................................................................................ 28
Oahu LiDAR QA Report May 16, 2014 Page 4 of 30
Executive Summary The following LiDAR quality assurance report documents Dewberry’s initial review of LiDAR data and derived products for the island of Oahu which was flown by Photo Science Inc (Quantum Spatial) for the NOAA Coastal Services Center. The data for Oahu covers approximately 602 square miles that consist of 694 tiles (1500 meters x 1500 meters). Each tile contains LAS point cloud data and a hydro-flattened DEM. Other deliverables include hydrographic breaklines and metadata. The figure below shows the locations of the project area (Figure 1)
Figure 1 - Location of the Oahu Project Area.
Oahu LiDAR QA Report May 16, 2014 Page 5 of 30
The LiDAR data and derived products were processed through Dewberry’s comprehensive quantitative/qualitative review. This multipart analysis determines the degree to which the data met expectations for completeness, relative accuracy, and conformity to specific project requirements for each data product. The LiDAR data for Oahu were thoroughly examined by Dewberry for completeness and conformity to project specifications. The data passes vertical accuracy requirements. However, there are a few qualitative issues that should be corrected by Photo Science Inc (PSI), including minor artifacts, eight (8) areas of aggressive misclassification, 110 areas of vegetation points remaining in ground, 47 flight line ridges, and vertical accuracy issues. Photo Science Inc. discussed the best approach to revising the vegetation artifacts within the project area and in areas where no ground was visible the best effort was made to include points that were likely close to the actual ground. The result is that in some area minor vegetation may still be present but was preferred over large areas of no data. All issues commented in the geodatabase and report have been resolved in a sufficient manner. There were no breakline issues identified in the review of the Oahu dataset. While 1.3% of the vertices on the breaklines were floating, a review of the hydro-flattened surface model identified that these were not impacting the DEM products. Dewberry did identify five (5) locations with minor water surface issues as well as two areas where there was a hard line between DEM tiles. Also, modifications made to the LiDAR and breakline datasets will require some DEMs to be re-processed. All water surface issues have been resolved in the updated deliverable. Project metadata and metadata for each deliverable was reviewed to verify it contained sufficient content. All metadata files were verified against the MetaParser (MP) tool to ensure the files meet Federal Geographic Data Committee (FGDC) standards. Dewberry found no MP errors but recommend several modifications to the metadata files to improve the quality and completeness of the content described and reported. Only one set of metadata was delivered for all four counties.
Contract:
Production Contractor:
Delivery #:
Dewberry
Recommendation:
Oahu LiDAR
Photo Science Inc (Quantum
Spatial) 2
It is Dewberry’s
recommendation to accept
the data
Data History:
� Initial Full Delivery of Project Area
� Final Revisions (only revised tiles redelivered)
Oahu LiDAR QA Report May 16, 2014 Page 6 of 30
DELIVERABLES SUMMARY FOR OAHU LIDAR
DELIVERABLE APPLICABLE ACCEPTANCE
CRITERIA DEWBERRY RECOMMENDATION
ALL-RETURN LAS POINT CLOUD
DATA
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 17, 18, 19, 23, 24, 25,
26
Accept
Accept with Comm ents
Return for Corrections
Reject
BREAKLINE GEODATABASE 11, 12, 13, 14, 15, 30, 31, 32, 33,
34
Accept
Accept with Comm ents
Return for Corrections
Reject
BARE EARTH DEMS 11, 12, 13, 14, 15, 19, 20, 21, 23,
24, 25, 26
Accept
Accept with Comm ents
Return for Corrections
Reject
LAS METADATA 22, 35
Accept
Accept with Comm ents
Return for Corrections
Reject
BREAKLINE METADATA 22, 35
Accept
Accept with Comm ents
Return for Corrections
Reject
DEM METADATA 22, 35
Accept
Accept with Comm ents
Return for Corrections
Reject
Oahu LiDAR QA Report May 16, 2014 Page 7 of 30
The applicable acceptance criteria refer to the numbered criteria found in “Appendix A-Acceptance Criteria.” The acceptance criteria were also outlined in the final Quality Plan created by Dewberry.
Overview The goal of this LiDAR QAQC Project is to provide high accuracy elevation datasets of multiple deliverable products including LiDAR, hydro-flattened digital elevation models (DEMs), and 3D breaklines for National Oceanic and Atmospheric Administration’s (NOAA) Coastal Services Center in partnership with the United States Geological Survey (USGS). Dewberry’s role is to provide Quality Assurance (QA) of the LiDAR data and supplemental deliverables provided by PSI that includes completeness checks, vertical accuracy testing, and a qualitative review of the bare earth surface. Each product is reviewed independently and against the other products to verify the degree to which the data meets expectations. This report documents the quality of the deliverables for the island of Oahu.
LiDAR Analysis The LiDAR data are reviewed on project, tile, and point level to determine the relative accuracy, proper classification and conformity to project requirements. This review begins with a computational analysis of the points for completeness and to determine point data format, projection, classification scheme, number of returns per pulse, and intensity values of the points.
LIDAR QUANTITATIVE REVIEW
One of the first steps in assessing the quality of the LiDAR is a vertical accuracy analysis of the ground models in comparison to surveyed checkpoints. NOAA CSC provided 100 checkpoints for the entire project area.
Survey Vertical Accuracy Checkpoints The following table lists all checkpoints surveyed for use in vertical accuracy testing.
Point ID
NAD83 UTM Zone 04N Ellipsoid
LAND COVER
Easting X (m) Northing Y (m) Elevation (m)
A1 619743.7113 2354269.44 18.507 Bare Earth
A3 619747.335 2354278.974 18.476 Bare Earth
AA1 618829.8053 2384173.359 17.727 Bare Earth
AB1 612010.3586 2391471.623 17.928 Bare Earth
AC1 632992.0205 2365916.694 19.554 Bare Earth
AD1 630339.1253 2372748.734 25.377 Bare Earth
ADA3 630441.4488 2372838.724 24.933 Bare Earth
ADB2 630211.2852 2372804.704 29.374 Bare Earth
B08 622613.376 2351916.275 17.75 Bare Earth
Oahu LiDAR QA Report May 16, 2014 Page 8 of 30
B1 622566.099 2351864.347 17.968 Bare Earth
B10 622581.681 2351915.549 17.912 Bare Earth
B11 622522.472 2351893.758 18.102 Bare Earth
B3 622550.1672 2351857.892 18.004 Bare Earth
C07 624789.266 2351337.281 34.144 Bare Earth
C1 624876.1261 2351379.508 34.051 Bare Earth
C2 624893.04 2351385.062 34.184 Bare Earth
D1 577888.466 2383748.136 20.973 Bare Earth
DB05 577862.9321 2383715.682 18.613 Bare Earth
F1 583143.3918 2374880.311 79.585 Bare Earth
G1 583327.2511 2372175.413 18.423 Bare Earth
GC05 583327.787 2372228.684 18.022 Bare Earth
I1 592896.0579 2355248.443 17.109 Bare Earth
IB01 592879.4312 2355207.89 16.858 Bare Earth
J1 599797.8668 2359894.065 30.795 Bare Earth
K1 604598.0909 2357388.098 16.869 Bare Earth
M1 586282.0668 2379593.416 829.533 Bare Earth
MA02 586228.7535 2379661.784 838.689 Bare Earth
MA04 586272.688 2379606.008 830.768 Bare Earth
MB01 586306.2008 2379593.342 826.809 Bare Earth
N1 592792.552 2388841.723 16.959 Bare Earth
O1 578941.1861 2386463.626 21.031 Bare Earth
R1 608309.6601 2365253.147 18.804 Bare Earth
RB02 608203.2828 2365171.242 18.72 Bare Earth
S1 603854.267 2363769.032 22.64 Bare Earth
SA01 603477.9427 2363669.805 27.226 Bare Earth
T1 602165.63 2373329.234 227.17 Bare Earth
TA01 602143.8106 2373355.009 227.383 Bare Earth
TB03 602235.6211 2373400.924 226.773 Bare Earth
U1 599301.239 2379193.727 314.586 Bare Earth
UA03 599239.7054 2379169.123 312.803 Bare Earth
VA03 617054.2118 2360579.022 61.155 Bare Earth
W1 610722.3246 2360145.184 22.395 Bare Earth
X1 630178.1982 2353564.88 18.039 Bare Earth
YA01 638340.249 2354750.735 20.025 Bare Earth
Z1 635492.5488 2359335.217 23.63 Bare Earth
ZA03 635427.0219 2359380.71 22.873 Bare Earth
A5 619813.0088 2354191.118 18.615 Forest
ACA1 632992.0566 2365890.7 19.493 Forest
B04 622578.721 2351855.2 17.926 Forest
B05 622590.351 2351868.615 17.805 Forest
B07 622608.512 2351893.843 17.713 Forest
B09 622641.181 2351935.317 17.471 Forest
B12 622467.89 2351838.997 17.788 Forest
B13 622473.561 2351790.54 17.984 Forest
C03 624876.488 2351399.376 34.444 Forest
C05 624822.168 2351387.563 34.921001 Forest
GC02 583396.929 2372213.874 18.305 Forest
Oahu LiDAR QA Report May 16, 2014 Page 9 of 30
GC04 583328.231 2372240.554 17.917 Forest
MA03 586222.3815 2379699.072 843.14 Forest
XA04 630220.2706 2353550.783 18.016 Forest
ACB2 632992.7596 2365972.18 19.543 Grass
DB03 577898.9438 2383718.369 20.244 Grass
SA02 603510.6584 2363695.422 30.699 Grass
SB03 603649.5037 2363825.666 25.478 Grass
UB01 599316.6744 2379158.071 314.683 Grass
Y1 638344.8681 2354743.135 20.368 Grass
YA02 638311.7903 2354760.386 20.075 Grass
DB02 577917.7307 2383760.655 21.592 Scrub/Shrub
IA04 592911.9227 2355187.494 17.008 Scrub/Shrub
KB02 604643.9905 2357410.574 17.24 Scrub/Shrub
OA01 578922.0158 2386458.629 20.685 Scrub/Shrub
RB03 608091.1242 2365172.939 17.728 Scrub/Shrub
UB02 599337.7647 2379120.878 313.926 Scrub/Shrub
WB03 610673.155 2360192.715 22.876 Scrub/Shrub
YA03 638251.5219 2354741.602 19.807 Scrub/Shrub
YA04 638255.7373 2354671.239 18.794 Scrub/Shrub
A2 619707.0055 2354225.296 18.845 Urban
A4 619471.0531 2354050.772 19.917 Urban
ACA2 633051.4336 2365944.213 19.661 Urban
ACB1 632972.0947 2365945.654 19.609 Urban
ADA1 630340.4396 2372772.207 24.339 Urban
AE1 623136.1632 2370677.572 19.532 Urban
B06 622613.405 2351872.943 18.029 Urban
C04 624855.165 2351447.324 36.452999 Urban
C06 624786.842 2351374.634 35.388999 Urban
DA06 577890.0955 2383779.567 21.83 Urban
FA01 583150.307 2374889.527 80.07 Urban
GC01 583388.169 2372190.408 18.56 Urban
GC03 583372.01 2372217.59 18.381 Urban
IA02 592883.5162 2355266.743 17.114 Urban
KB01 604626.0241 2357396.197 16.864 Urban
MA01 586265.8412 2379604.132 830.947 Urban
MB02 586340.8071 2379573.762 825.326 Urban
RB01 608281.4871 2365261.562 18.994 Urban
SB02 603678.2132 2363796.412 24.716 Urban
TA04 602208.5594 2373344.286 226.749 Urban
TB01 602155.6261 2373318.403 226.467 Urban
V1 617042.5117 2360456.316 57.636 Urban
WB01 610701.5584 2360164.045 22.278 Urban
XA01 630166.3778 2353588.656 19.132 Urban
Oahu LiDAR QA Report May 16, 2014 Page 10 of 30
Table 1: Oahu LiDAR surveyed accuracy checkpoints
Figure 3 – Checkpoint distribution for Oahu LiDAR data.
The vertical accuracy assessment compares the measured survey checkpoint elevations with those of the TIN as generated from the bare-earth LiDAR. The X/Y locations of the survey checkpoints are overlaid on the TIN and the interpolated Z values of the LiDAR are recorded. These interpolated Z values are then compared with the survey checkpoint Z values and this difference represents the amount of error between the measurements. Once all the Z values are recorded, the FVA, CVA, and SVA are computed. The data were analyzed by Dewberry to assess the accuracy of the data. The review process examined the various accuracy parameters as defined by the scope of work. The overall descriptive statistics of each dataset were computed to assess any trends or anomalies. FVA (Fundamental Vertical Accuracy) is determined with check points located only in the open terrain (grass, dirt, sand, and/or rocks) land cover category, where there is a very high probability that the LiDAR sensor will have detected the bare-earth ground surface and where random errors are expected to follow a normal error distribution. The FVA determines how well the calibrated LiDAR sensor performed. With a normal error distribution, the vertical accuracy at the 95% confidence level is computed as the vertical root mean square error (RMSEz) of the checkpoints x 1.9600. For the Oahu LiDAR project, vertical accuracy must be 24.5 cm or less based on an RMSEz of 12.5 cm x 1.9600.
CVA (Consolidated Vertical Accuracy) is determined with all checkpoints in all land cover categories combined where there is a possibility that the LiDAR sensor and post-processing may
Oahu LiDAR QA Report May 16, 2014 Page 11 of 30
yield elevation errors that do not follow a normal error distribution. CVA at the 95% confidence level equals the 95th percentile error for all checkpoints in all land cover categories combined. The Oahu LiDAR Project CVA standard is 36.0 cm at the 95% confidence level. The CVA is accompanied by a listing of the 5% outliers that are larger than the 95th percentile used to compute the CVA; these are always the largest outliers that may depart from a normal error distribution. Here, Accuracyz differs from CVA because Accuracyz assumes elevation errors follow a normal error distribution where RMSE procedures are valid, whereas CVA assumes LiDAR errors may not follow a normal error distribution in vegetated categories, making the RMSE process invalid. SVA (Supplemental Vertical Accuracy) is determined for each land cover category other than open terrain. SVA at the 95% confidence level equals the 95th percentile error for all checkpoints in each land cover category. The Oahu LiDAR Project SVA target is 36.0 cm at the 95% confidence level. Target specifications are given for SVA’s as individual land cover categories may exceed this target value as long as the overall CVA is within specified tolerances. Again, Accuracyz differs from SVA because Accuracyz assumes elevation errors follow a normal error distribution where RMSE procedures are valid, whereas SVA assumes LiDAR errors may not follow a normal error distribution in vegetated categories, making the RMSE process invalid. The relevant testing criteria are summarized in Table 2.
Quantitative Criteria Measure of Acceptability
Fundamental Vertical Accuracy (FVA) in open terrain only using RMSEz *1.9600
0.245 meters (based on RMSEz (0.125m) * 1.9600)
Consolidated Vertical Accuracy (CVA) in all land cover categories combined at the 95% confidence level
0.360 meters (based on combined 95th percentile)
Supplemental Vertical Accuracy (SVA) in each land cover category separately at the 95% confidence level
0.360 meters (based on 95th percentile for each land cover category)
Table 2 ― Acceptance Criteria
Vertical Accuracy Results – Swath LiDAR Table 3 outlines the calculated RMSEz and associated statistics, in meters, of the swath data as required by USGS V1 specificaitons.
Swath Accuracy Results
100 % of
Totals
# of Points
RMSEz (m) Open
Terrain Spec=0.0925
m
FVA- Fundamental
Vertical Accuracy
((RMSEz x 1.9600)
Spec=0.181 m
Mean (m)
Median (m)
Skew Std Dev (m)
Min (m)
Max (m)
Open Terrain 46 0.10 0.20 -0.01 -0.01 -0.32 0.10 -0.25 0.21
Oahu LiDAR QA Report May 16, 2014 Page 12 of 30
Table 3 - The table shows the calculated RMSEz values, in meters, as well as associated statistics of the errors for the LiDAR Swaths in the Oahu project area.
Vertical Accuracy Results – Classified LiDAR Table 4 outlines the calculated RMSEz and associated statistics, in meters, while
LiDAR Vertical Accuracy Results
Land Cover Category # of Points
FVA ― Fundamental
Vertical Accuracy (RMSEz x 1.9600)
Spec=0.245 m
CVA ― Consolidated
Vertical Accuracy (95th
Percentile) Spec=0.36 m
SVA ― Supplemental
Vertical Accuracy (95th
Percentile) Target=0.36 m
Consolidated 100 0.268 Bare Earth-Open
Terrain 46 0.198
Urban 24 0.250
Combined Vegetation 30 0.332
Table 5 outlines vertical accuracy as computed by the different methods, in meters.
Descriptive Statistics
100 % of Totals
# of Points
RMSEz (m)
Open Terrain
Spec=0.125 m
Mean (m)
Median (m)
Skew Std Dev (m)
Kurtosis Min (m)
Max (m)
Consolidated 100 -0.017 -0.019 0.869 0.126 3.005 -
0.290 0.507
Open Terrain 46 0.101 -0.007 -0.004 -0.482 0.102 0.504
-0.267 0.202
Urban 24 -0.056 -0.084 0.970 0.117 2.912 -
0.290 0.303
Tall Weeds and Crops 7 -0.018 0.022 -0.296 0.090 -2.523
-0.129 0.076
Brush Lands and Trees 9 0.128 0.107 0.019 0.224 0.870
-0.275 0.507
Forested and Fully Grown 14 -0.071 -0.075 -0.592 0.075 1.257
-0.244 0.060
Combined Vegetation 30 0.001 -0.018 1.306 0.160 3.012
-0.275 0.507
Table 4 - The table shows the calculated RMSEz values, in meters, as well as associated statistics of the errors for the Oahu project area for the final classified LiDAR deliverable.
LiDAR Vertical Accuracy Results
Oahu LiDAR QA Report May 16, 2014 Page 13 of 30
Land Cover Category # of Points
FVA ― Fundamental
Vertical Accuracy (RMSEz x 1.9600)
Spec=0.245 m
CVA ― Consolidated
Vertical Accuracy (95th
Percentile) Spec=0.36 m
SVA ― Supplemental
Vertical Accuracy (95th
Percentile) Target=0.36 m
Consolidated 100 0.268 Bare Earth-Open
Terrain 46 0.198
Urban 24 0.250
Combined Vegetation 30 0.332
Table 5 - The table shows the calculated FVA, CVA, and SVA, in feet, at the 95% confidence level.
Table 6 lists the 5% outliers that are larger than the 95th percentile, or 0.616 feet.
Point ID
NAD83 UTM Zone 04N Ellipsoid
LiDAR Z
(m) Delta Z AbsDeltaZ
Easting X (m) Northing Y (m) Survey Z
(m)
E1 579828.0295 2380581.60 18.946 19.8359 0.890 0.890
JA01 599809.7605 2359896.26 32.291 30.9508 -1.340 1.340
ABA2 612038.1058 2391490.663 15.835 17.73010 1.895 1.895
ZA04 635459.3668 2359482.417 22.510999 20.53620 -1.975 1.975
AEA3 623266.8992 2370776.838 19.722 17.04290 -2.679 2.679
Table 6 ― 5% Outliers
Figure 4 illustrates the magnitude of the differences between the QA/QC checkpoints and LiDAR data.
Oahu LiDAR QA Report May 16, 2014 Page 14 of 30
Figure 4 – Magnitude of Elevation Discrepancies
LIDAR COMPLETENESS REVIEW
Dewberry received 637 LiDAR tiles for the project area. The LiDAR was delivered in LAS format 1.2, point data format 1 is used, and all data have intensity values. The LAS tiles are named appropriately according to the SOW and have correct extents (1500 m x 1500 m). The majority of all LiDAR tiles were within the specified combined scan angle of 40°. However, the maximum combined scan angle is 54°, but Dewberry has reviewed all tiles for relative accuracy and flight line ridge issues. No issues were identified that could be attributed to the high scan angles. All spatial projection information was correct and is as follows:
� Horizontal Datum: NAD83 (PA11) � Vertical Datum: Ellipsoid � Projection: UTM 04N � Horizontal and Vertical Units: Meters
Each record includes the following fields (among others):
� X, Y, Z coordinates � Flight line data � Intensity value � Return number
-0.400
-0.300
-0.200
-0.100
0.000
0.100
0.200
0.300
0.400
0.500
0.600M
ete
rs
Checkpoint Errors
Open Terrain
Urban
Combined
Vegetation
Oahu LiDAR QA Report May 16, 2014 Page 15 of 30
� Number of returns � Scan direction � Edge of flight line � Scan angle � Classification � Adjusted GPS time
The LiDAR data has been classified to contain the following classes:
Required Classes � Class 1 (Unclassified) � Class 2 (Bare Earth) � Class 7 (Low Points and Outliers) � Class 9 (Water) � Class 10 (Ignored Ground) � Class 17 (Unclassified Overlap) � Class 18 (Filtered Bare-Earth Overlap) � Class 25 (Water Overlap)
All tiles met the project requirement to have 20% overlap on adjoining swaths.
Point Count/Elevation Analysis To verify the content of the data and validate the data integrity, a statistical analysis was performed on each tile. This process allows Dewberry to review 100% of the data at a macro level to identify any gross outliers. The statistical analysis consists of first extracting the header information and then reading the actual records and computing the number of points, minimum, maximum, and mean elevation for each class. Minimum and maximum for other relevant variables are also evaluated. No major anomalies were identified. Each tile was queried to extract the number of LiDAR points. The required nominal point spacing for the project is 1.0 meter. By utilizing the full point cloud, the Oahu project area was determined to have a nominal point spacing of greater than 1 point per square meter which satisfies the project requirements.
LIDAR QUALITATIVE REVIEW
The goal of Dewberry’s qualitative review is to assess the continuity and the level of cleanliness of the bare earth product. Each LiDAR tile is expected to meet the following acceptance criteria:
� The point density is homogenous and sufficient to meet the user’s needs; � The ground point have been correctly classified (no man-made structures or
vegetation remains, no gaps except over water bodies); � The ground surface model exhibits a correct definition (no aggressive classification,
no over-smoothing, no inconsistency in the post-processing); � No obvious anomalies due to sensor malfunction or systematic processing artifacts
are present (data voids, spikes, divots, ridges between flight lines or tiles, cornrows, etc);
� Residual artifacts <5%
Oahu LiDAR QA Report May 16, 2014 Page 16 of 30
Dewberry analysts performed a visual inspection of 100% of the bare earth data digital terrain model (DTM). 100% of the project data was reviewed at the micro and macro levels. The DTMs are built by first creating a fishnet grid of the LiDAR masspoints with a grid distance equal to or better than the final DEM deliverables. Then a triangulated irregular network is built based on this gridded DTM and displayed as a 3D surface. A shaded relief effect was applied which enhances 3D rendering. Quick Terrain Modeler, the software used for visualization allows the user to navigate, zoom and rotate models and to display elevation information with an adaptive color coding in order to better identify anomalies. Models can also be viewed by point density, in which areas meeting the specified point density threshold are displayed green and areas not meeting the point density threshold are displayed red. This can help to identify void areas and areas that are misclassified. As the surface model is created from ground only points, sparse or red areas are expected over buildings, water, and dense vegetation where there is poor LiDAR penetration. The table below shows a breakdown of the calls made during the first review of the project data.
Issue Number of Occurrences Remaining Issues
Aggressive Misclassification 9 0
Artifacts 5 0
Divots 1 0
Corn Rows 1 0
Flight Line Ridges 51 0
Vegetation Artifacts in Ground
129 0
Voids 7 0
Total 203 0
Table 7 – Breakdown of LiDAR qualitative edit calls.
Aggressive Misclassification Aggressive misclassification calls in this document imply that LiDAR points are unclassified in the delivered dataset when they should be classified to ground. This call indicates areas where some class 1 points could be reclassified to class 2, ground, to improve detail in the surface model and to more correctly model surface features. There are nine (9) instances of aggressive misclassification identified in this project area. An example of aggressive misclassification edit calls is found below and a geodatabase that contains the locations of all edit calls accompanies this report. The majority of these calls have been placed on the tops of ridges where points are present that would represent the peaks.
All areas marked for revisions have been addressed. While some minor misclassification still exists on the very tops of ridges they are within the 2% requirement for incorrectly classified data.
Oahu LiDAR QA Report May 16, 2014 Page 17 of 30
Figure 5 – LAS Tile 4QFJ619380. The top image shows a view of the terrain with edit call highlighted. The bottom image shows a profile view of the LAS cloud colored by classification (ground-purple,
unclassified-yellow) is shown on the bottom. Valid unclassified points should be reclassified to ground to improve the definition of the bare-earth surface. The image on the right shows the
corrected area with additional point being classified to ground.
Artifacts Artifacts are features that are left in the ground model that should be removed. There are five (5) artifacts identified in the project area and include bridges and structures. Vegetation has been separated out into a separate class for this project as the majority of issues were related to vegetation artifacts. These should be removed in order to improve the bare-earth surface model. Examples of artifact edit calls can be found below and a geodatabase that contains the locations of all edit calls accompanies this report.
All artifacts have been sufficient resolved.
Oahu LiDAR QA Report May 16, 2014 Page 18 of 30
Figure 6 - Tile 4QFJ612356 Profile view of the LAS cloud colored by classification (ground-pink) is shown on top while bare-earth TIN colored by elevation is shown on bottom. The features can be
seen protruding from the ground and should be reclassified to unclassified (class 1). The area around the airport was highlighted for review of artifacts as a number of them were present on the
surface. The bottom surface model shows all of these features have been removed from the ground.
Oahu LiDAR QA Report May 16, 2014 Page 19 of 30
Corn Rows One area within the project area exists where corn rowing is prevalent throughout the area and is visible at the review scale. The extents of this area have been outlined in the LAS edit calls polygon. The edges of this area are also visible with a sharp change in elevation and overall LiDAR smoothness.
The issue regarding corn rows has been resolved and is no longer present in the dataset.
Figure 7 – Tile 4QFJ624365. Example show visible corn row effect throughout this tile and surrounding area. The bottom image shows that the corn rows have been corrected.
Oahu LiDAR QA Report May 16, 2014 Page 20 of 30
Vegetation Artifacts in Ground Vegetation artifact calls in this document imply that LiDAR points are incorrectly classified in the delivered dataset and should be re-classified to class 1 (unclassified). There are 123 instances of vegetation artifacts identified in this project area. Because this issue was prevalent in the dataset polygon edit calls were placed in areas where the issue was most visible.
All vegetation artifacts have been sufficiently addressed. In some cases minor vegetation does remain in the ground but based on input from NOAA CSC the limited vegetation was acceptable and does not negatively impact the use of the data.
Figure 8 & 9 – Tile 4QFK606400. The top image is a profile view of the LAS cloud colored by classification (unclassified-yellow, ground - pink). In this example ground points start to move to the
lower extent of the vegetation and should be reclassified as class 1. The second images show the corrected area.
Oahu LiDAR QA Report May 16, 2014 Page 21 of 30
LIDAR RECOMMENDATION
Dewberry recommends the LiDAR be accepted by NOAA CSC.
Breakline Analysis A qualitative/quantitative review was completed on the project area breaklines. The comprehensive review consisted of a visual review of the breaklines for completeness in compilation and horizontal placement as well as proper feature coding. This visual analysis was followed by several automated tests for hydro-flattening and topology using ESRI Data Reviewer tools and proprietary tools developed by Dewberry. The breakline review followed the Breakline QA/QC Checklist provided in Appendix B. This QA/QC checklist was also provided in Dewberry’s Quality Plan.
BREAKLINE DATA OVERVIEW
The Breakline qualitative review starts with an overview. First, the ESRI geodatabase is reviewed in ArcCatalog for correct spatial projections, data organization, and to ensure all necessary feature classes are present and are properly populated. The delivered geodatabases contained the correct feature classes, shown below:
� TIDAL_WATERS � STREAMS_AND_RIVERS � PONDS_AND_LAKES
The coordinate system of the delivered breaklines is correct and is as defined below:
� Horizontal Datum: NAD83 (PA11) � Vertical Datum: Ellipsoid � Projection: UTM 04N � Horizontal and Vertical Units: Meters
BREAKLINE QUALITATIVE REVIEW
The breakline qualitative review includes reviewing data for completeness, validating the horizontal placement of breaklines, and verifying the coding and attribution of breaklines.
The breaklines were reviewed against intensity imagery Dewberry creates for its QC process. A review was performed on 100% of the data in an ESRI environment to validate data collection consistency and to validate all necessary features were collected. There were no edit calls requiring revisions by Photo Science to the breakline dataset.
BREAKLINE QUANTITATIVE REVIEW
The Quantitative Vertical Analysis compares the breakline vertices against the bare-earth LiDAR data. Dewberry begins this process by converting all breakline vertices to points. At the same time an ESRI Terrain is created from the LiDAR using ground and water points. The LiDAR elevation, extracted from the terrain, is recorded for every breakline vertex. An analysis of the differences in elevation between the breakline vertices and LiDAR is conducted to determine the vertical accuracy of the breakline collection.
Oahu LiDAR QA Report May 16, 2014 Page 22 of 30
Dewberry found no issues in this portion of the review.
TOPOLOGY
One of the requirements of hydro breaklines intended for modeling is valid topology. Dewberry tested the topology using ESRI’s Data Reviewer extension and proprietary tools to ensure that the breakline vertices are snapped together, that hydro-lines fulfill monotonicity requirements within a specified tolerance, that all water bodies are flat within a tolerance, and that all breaklines have elevations defined. These data checks allow automated validation of 100% of the data and are used on every delivery of data to ensure consistent data integrity. The data checks used are listed in detail in Appendix B-The Breakline QA/QC Checklist. The Breakline QA/QC Checklist was also provided in Dewberry’s Quality Plan. Dewberry found no issues in this portion of the review.
Monotonicity All stream/river breaklines should show continuous downhill flow. Some exceptions may occur at confluences of rivers or where tributaries or other off-shoots of a river exist. Monotonic breaklines are necessary to produce hydro-flattened DEMs. Dewberry found no issues in this portion of the review.
Connectivity Connectivity, or different Z at intersections, ensures that all overlapping vertices that have the same X-Y coordinates also have the same Z-elevations. While some exceptions can occur for vertical structures, such as sea walls or dams, most overlapping vertices should have the same Z-elevation. This means that the first and last vertices for a closed feature should have the same elevation and that vertices from one feature that connect to vertices from another feature, such as a stream that drains into a water body, should have the same elevation. Connectivity issues can cause false waterfalls within a feature or between two features. Dewberry found no issues in this portion of the review.
BREAKLINE RECOMMENDATION
Dewberry recommends accepting the breakline data as delivered
Hydro-flattened Digital Elevation Model Analysis Dewberry received 637 hydro-flattened bare earth DEMs as part of the deliverables for the project area. Dewberry used proprietary scripts and tools to ensure all DEMs have the correct formatting, cell size, projection, and extents. Dewberry used ESRI ArcMap and Global Mapper software to review all DEMs for completeness and qualitative analysis.
OVERVIEW
Dewberry ran proprietary tools on all delivered DEMs to verify formatting, cell size, extents, and projection information. All DEMs were correctly formatted:
� DEM type: IMG � Cell Size: 1 meter � Extents: 1500 m x 1500 m tiles
Oahu LiDAR QA Report May 16, 2014 Page 23 of 30
The coordinate system of the delivered DEMs is correct and is as defined below:
� Horizontal Datum: NAD83 (PA11) � Projection: UTM 04N � Horizontal Units: Meters
DEM QUANTITATIVE REVIEW
Vertical Accuracy Results – Digital Elevation Model (DEM) Table 8 outlines the calculated RMSEz and associated statistics, in meters, while
LiDAR Vertical Accuracy Results
Land Cover Category # of Points
FVA ― Fundamental
Vertical Accuracy (RMSEz x 1.9600)
Spec=0.245 m
CVA ― Consolidated
Vertical Accuracy (95th
Percentile) Spec=0.36 m
SVA ― Supplemental
Vertical Accuracy (95th
Percentile) Target=0.36 m
Consolidated 100 0.268 Bare Earth-Open
Terrain 46 0.198
Urban 24 0.250
Combined Vegetation 30 0.332
Table 9 outlines vertical accuracy as computed by the different methods, in meters.
Descriptive Statistics
100 % of Totals
# of Points
RMSEz (m)
Open Terrain
Spec=0.125 m
Mean (m)
Median (m)
Skew Std Dev (m)
Kurtosis Min (m)
Max (m)
Consolidated 100 -0.017 -0.019 0.869 0.126 3.005 -
0.290 0.507
Open Terrain 46 0.101 -0.007 -0.004 -0.482 0.102 0.504
-0.267 0.202
Urban 24 -0.056 -0.084 0.970 0.117 2.912 -
0.290 0.303
Tall Weeds and Crops 7 -0.018 0.022 -0.296 0.090 -2.523
-0.129 0.076
Brush Lands and Trees 9 0.128 0.107 0.019 0.224 0.870
-0.275 0.507
Forested and Fully Grown 14 -0.071 -0.075 -0.592 0.075 1.257
-0.244 0.060
Combined Vegetation 30 0.001 -0.018 1.306 0.160 3.012
-0.275 0.507
Oahu LiDAR QA Report May 16, 2014 Page 24 of 30
Table 8 - The table shows the calculated RMSEz values, in meters, as well as associated statistics of the errors for the Oahu project area for the final DEM deliverable.
LiDAR Vertical Accuracy Results
Land Cover Category # of Points
FVA ― Fundamental
Vertical Accuracy (RMSEz x 1.9600)
Spec=0.245 m
CVA ― Consolidated
Vertical Accuracy (95th
Percentile) Spec=0.36 m
SVA ― Supplemental
Vertical Accuracy (95th
Percentile) Target=0.36 m
Consolidated 100 0.268 Bare Earth-Open
Terrain 46 0.198
Urban 24 0.250
Combined Vegetation 30 0.332
Table 9 - The table shows the calculated FVA, CVA, and SVA, in feet, at the 95% confidence level.
QUALITATIVE REVIEW
Dewberry performed a visual analysis on 100% of the delivered DEMs. The DEMs were reviewed in Global Mapper or ESRI ArcMap software. The DEMs were reviewed with hillshades, which allow the viewer to see the DEMs as if in 3D. This helps with the identification of issues and anomalies. The DEM is required to be free of artifacts, gaps, and artificial smoothing. A breakdown of the edit calls made during the review can be seen in the table below.
Issue Number of Occurrences Remaining Issue
Artifacts 7 0
Total 7 0
Table 6 – Breakdown of DEM qualitative edit calls
Artifacts Dewberry found seven tiles that contained minor artifacts in the DEM surface likely from processing either within a single tile or between two adjacent tiles. All artifacts have been addressed in the redelivery of the dataset.
DEM RECOMMENDATION
It is Dewberry’s recommendation that the DEMs be accepted by NOAA CSC.
Oahu LiDAR QA Report May 16, 2014 Page 25 of 30
Metadata Photo Science delivered five project level metadata files, in XML format for the raw swaths, final classified LAS, breaklines, and DEMs. As well as a single project level metadata file. Additionally, metadata was provided for each lift as required by USGS. Dewberry reviewed the metadata files for correct formatting and for sufficient content. All metadata files meet FGDC standards and were deemed error free by the MetaParser (MP) tool developed by the United States Geological Survey.
All metadata files have similar abstracts. Each lineage step has the beginning and end dates of acquisition and provides detailed information for the acquisition and processing of the LiDAR and the breakline collection process. However, the LiDAR metadata should include references to the additional overlap classes that were used in the LiDAR dataset.
The metadata files have been updated based on the recommendations.
METADATA RECOMMENDATION
Dewberry recommends the metadata be accepted by NOAA CSC.
Other Comments Along with this report, Dewberry is providing a GDB named Oahu_D1_QC_20140321.gdb” that contains all LiDAR, breakline, and DEM edit calls from our review of the Oahu deliverable. PSI should comment for each call whether or not the issue was corrected. If no corrections are made, the reason why should be documented in the feature class. This GDB will provide a record of the review, modification, and response process used on the Oahu data.
Recommendations Summary The following represents a summary of Dewberry’s recommendations for Photo Science Inc (Quantum Spatial). These recommendations can be found throughout the various sections of this report but are summarized here for convenience. LIDAR:
1. Recommended for acceptance.
BREAKLINES: 1. Recommended for acceptance.
DEMS: 1. Recommended for acceptance.
METADATA: 1. Recommended for acceptance.
Oahu LiDAR QA Report May 16, 2014 Page 26 of 30
APPENDIX A ─ ACCEPTANCE CRITERIA Criteria Tested Characteristic Measure of Acceptability
LiDAR Data Acquisition Acceptance Criteria
1. Returns per pulse LiDAR sensor shall be capable of recording up to 3 (or more) returns per pulse, including 1st and last returns
2. Scan angle ≤ ±20 degrees on each side of nadir, i.e., maximum Field of View = 40 degrees
3. Swath overlap Nominal sidelap of 15% on adjoining swaths. Any data gaps between the geometrically usable portions of the swaths will be rejected.
4. Design pulse density (nominal)
≥ 1 pulse/m2 per swath; assessment to be made against single swath, first return located within the geometrically usable center portion of each swath.
5. GPS procedures Base stations for GPS surveys shall be based on first or second order survey control stations that are part of the National Geodetic Survey’s National Spatial Reference System (NSRS). New control stations will be sufficiently monumented to hold their position.
6. Collection Conditions Ground is snow free unless approved and documented by NOAA and its partners Streams must be within their banks
LiDAR Accuracy Acceptance Criteria
7. Vertical Accuracy Fundamental Vertical Accuracy (FVA) 12.5cm RMSEz or 24.5 cm at the 95% confidence level Consolidated Vertical Accuracy (CVA) for 3 land cover classes that are TBD, should have an Accuracyz of 36cm There shall be minimal vertical offset (7cm RMSE) between adjacent flight lines
8. Horizontal Accuracy RMSExy shall ≤ 4.0’
Geographic Coverage and Continuity Acceptance Criteria
9. Coverage No voids between swaths. No voids because of cloud cover or instrument failure. Voids within a single swath ≥ (4*NPS)2 will not be acceptable except for voids caused by water bodies, low reflectivity or where appropriately filled-in by an overlapping swath.
10. Aggregate 1st return density
Barring non-scattering areas (e.g., open water, wet asphalt); no voids of more than 4*NPS^2; point spacing (1st return) within each swath must have an average of 1 m, not including overlap points. Acceptable data voids identified in Acceptance Criterion #9 above are excluded from this requirement.
Spatial Reference Framework
11. Vertical Datum NAVD 88, processed with Geoid12 – Gridded products shall be Local Mean Sea Level as defined by the NOAA tide station in Honolulu Harbor (#1612340)
12. Horizontal Datum NAD 83 (PA11)
13. Projection UTM Zone 4
14. Vertical Units Meters (orthometric heights), 3 decimal places
15. Horizontal Units Meters, 3 decimal places
Deliverables
16. Report of Survey Text report that describes survey methods; x,y,z results; file formats; file naming schemes; tiling schemes., .pdf, .doc, or .odt format. The survey data and report shall be delivered on the same media as the actual data.
continued
Oahu LiDAR QA Report May 16, 2014 Page 27 of 30
Criteria Tested Characteristic Measure of Acceptability
17. Flight Lines as-flown ESRI geodatabase format (vector) with attribution for start and stop time as well as dates. Dates must be in Julian date format with times generated using either a 12 or 24 hour cycle. GPS time is not acceptable.
18. All-return point cloud List of all valid returns in LAS 1.2 format. For each return: GPS week, GPS second, easting, northing, elevation, intensity, return#, return classification. May include additional attributes that are outlined in the LiDAR SOW. No duplicate entries. GPS second shall be reported to the nearest microsecond (or better). Easting, northing, and elevation shall be reported to nearest 0.01 m (nearest 0.01 ft). Classification of returns shall be as complete as is feasible (including classes: 1. Unclassified, 2. Ground, 7. Noise, 9. Water, 10. Breakline Proximity points, 17. Unclassified overlap, 18. Ground overlap, 25. Water overlap) without avoidable return misclassification.
19. File naming convention LAS Files -1500m x 1500m on even boundaries. Naming shall conform to nomenclature provided by NOAA and its partners.
20. Digital Elevation Model (DEM) of bare-earth w/ breaklines
DEM of bare-earth terrain surface (1 meter grid), interpolated using a triangulated irregular network (TIN) from identified ground points and hydro-flattened, 2.5 and 3D breaklines. DEM deliverables will be in IMG format
21. DEM Artifacts DEM shall have no tiling artifacts, no gaps and no artificial smoothing at tile boundaries. Areas outside survey boundary shall be coded as NoData. Internal voids (e.g., open water areas) may be coded as NoData.
22. Formal metadata See SOW instructions on formal metadata
23. Ground Points (Bare Earth)
Post-processed to remove structures and vegetation with <5% residual artifacts
24. Inconsistent Post-Processing, Editing
No visible variations in LiDAR data caused by alternating processing techniques. Bare earth surface must be consistent in both the removal of features and the features left in the ground.
25. Over-Smoothing Smoothing techniques shall not remove topographic features necessary to define drainage structures.
26. LAS Artifacts No obvious artifacts, spikes, holes or blunders; no cornrows
Usability Acceptance Criteria
27. Internal file formats Files shall have consistent internal formats
28. Compressed files Files shall not be compressed, unless documented by NOAA and partners
29. Ancillary geographic feature data
Ancillary geographic feature data represented as vector data types shall have complete and correct associated projection files.
Breakline Acceptance Criteria
30. Completeness 3D Breaklines collected for: • Streams and Rivers • Ponds and Lakes • Tidal Waters
31. Monotonicity • Double Line Streams shall generally maintain a consistent down-hill flow and be collected in the direction of flow – some natural exceptions will be allowed.
32. Vertical Consistency • Closed Waterbodies shall maintain a constant elevation at all vertices. • Vertices should not have excessive min or max z-values when compared to adjacent vertices •Vertical variance between breaklines & LiDAR DTM: Continuous segments shall not float above existing terrain >0.3 ft. Features should not be excessively buried compared to surrounding terrain. • Intersecting features shall maintain connectivity in X,Y,Z planes
Oahu LiDAR QA Report May 16, 2014 Page 28 of 30
Criteria Tested Characteristic Measure of Acceptability
33. Topology • Streams and Rivers must not self-intersect or intersect • Lakes and Ponds must not overlap or have gaps • Features must not have unnecessary dangles or boundaries
34. Horizontal Placement Breaklines are required to meet a tolerance of 2x the nominal post spacing for the horizontal placement when breaklines are compared to intensity imagery. The requirement applies only to the inside placement of the hydrographic feature, or inside the shoreline. If the feature intersects the bank of the feature so that ground is captured as water, it will still be considered an error.
35. Metadata Metadata must be FGDC compliant and contain sufficient detail to document source materials, projections, datums, processing steps, etc.
Oahu LiDAR QA Report May 16, 2014 Page 29 of 30
Appendix B- QA/QC Checklists LiDAR Checklist Overview
Correct number of files is delivered and all files are in LAS format 1.2 - Passed LiDAR Swaths have been delivered and are in LAS format 1.2 – Passed LAS statistics have been run
Quantitative Review/Accuracy Assessment Absolute and relative accuracy of data (vertical and horizontal) should be verified prior to classification and subsequent product development
Control Points and Checkpoints in ASCII format FVA: RMSEz = 12.5cm or ACCURACYz at 95% = 24.5 cm CVA: ACCURACYz at 95% =36 cm Horizontal Accuracy: 4’ at 95% Final accuracy report has been produced
Project Requirements for LAS
Nominal Pulse spacing no greater than 1 meter, assessment made against single swath- Passed LiDAR is free from data voids, 4m2, except for over water bodies, areas of low NIR - Passed
reflectivity such as asphalt or composition roofing, and where appropriately filled-in by another swath Distribution of points is uniform and free of clustering: A rectangular grid with 1m
cell size overlaid on data shows at least 90% of the cells in the grid contain at least 1 LiDAR point when tested against 1st return only data- Passed
Scan angle does not exceed 40° Passed Flightline overlap of 20% or greater- Passed Data collected to cover project area shapefiles- Passed. Project does not cover full extent but has been
approved by Client. LAS header includes all required fields according to the ASPRS LAS 1.2 specifications- Passed with some
minor inconsistencies in the GUID files LAS files include the mandatory GeoKey DirectoryTag variable length header - Passed Point Data Record Format 1 is used - Passed Intensity field is required - Passed Classification is as follows: Class 1-Unclassified, Class 2-Ground, Class 7-Low Points/Noise, Class 9-Water,
Class 11 - withheld. Within any 1km x 1km area, no more than 2% of points will possess an erroneous classification value. – Passed, also includes classes 10, 17, 18, 25.
LiDAR data is referenced to NAD83 (PA11), UTM Zone 4 and NAVD88, using latest GEOID model (GEOID12). All units shall be in meters to 2 decimal places – Passed. Data was submitted in ellipsoid heights but that has been accepted by Client.
LiDAR data gaps between adjacent flight lines larger than 2 meters are not acceptable, but overlap points will be supplied. - Passed
Bridge removal is consistent across the project area following the bridge collection guidelines- Passed. One edit call has been placed for a single bridge to be removed.
Qualitative Review
LiDAR tiles reviewed for errors, anomalies, and incorrect project specifications – Passed Bare earth surface has been consistently classified, processed, and edited - Passed Profiles should be used to confirm every potential issue can be modified or corrected prior to marking it with
an edit call. Dewberry’s LiDAR SOP and standardized calls should be used during QC. - Passed If tiles are a re-submittal, corrections are verified to be acceptable Passed
Metadata
Project level LiDAR metadata XML file is error free as determined by the USGS MP tool Metadata content contains sufficient detail and all pertinent information regarding source materials,
projections, datums, processing steps, etc. – Passed
Breakline Checklist Overview
All Feature Classes are present in GDB All features have been loaded into the geodatabase correctly.
Oahu LiDAR QA Report May 16, 2014 Page 30 of 30
The breakline topology inside of the geodatabase has been validated (No Self-Intersection, No Overlap/Gaps, No Dangles).
The correct project level metadata is provided for each feature class within the geodatabase – Passed If geodatabase is a re-submittal, verify all edit calls placed in the QAQC database have been corrected and
comment fields are filled out correctly. Projection/coordinate system of GDB is accurate with project specifications (NAD83 – PA11, UTM Zone 4
and Ellipsoid Heights). All units in meters to 2 decimal places Perform Completeness check on breaklines using intensity imagery
Check entire dataset for missing features that were not captured, but should be to meet baseline specifications or for consistency
Breaklines are compiled to correct tile grid boundary and there is full coverage without overlap Breaklines are correctly edge-matched to adjoining datasets.
Compare Breakline Z elevations to LiDAR elevations
Compare all breakline vertex elevations to the LiDAR to ensure they are within the project tolerance. The impact the breaklines have on the final DEMs should be the deciding factor as to if there is an issue that requires corrections. Delta Z differences caused by noise within the LiDAR is expected and are not considered errors. – Some floating and digging vertices exist in the dataset. However, the impact on the DEM was minimal and any changes to the elevations would have caused the majority of the breakline to either dig or float greater than the existing feature.
Perform automated data checks using PLTS The following data checks are performed utilizing ESRI’s PLTS extension. These checks allow automated validation of 100% of the data. Error records can either be written to a table for future correction, or browsed for immediate correction. PLTS checks should always be performed on the full dataset.
Perform “adjacent vertex elevation change check” on the Waterbody feature class (Elevation Difference Tolerance=.001 feet).
Perform “unnecessary polygon boundaries check” on Waterbody and Coastal shoreline feature classes. Perform “different Z-Value at intersection check” Perform “duplicate geometry check” Perform “geometry on geometry check” Perform “polygon overlap/gap is sliver check”
Perform Dewberry Proprietary Tool Checks
Perform monotonicity check on stream banks to ensure consistent downhill flow on hydrographic features Perform connectivity check for intersecting or touching features
Metadata
Each XML file (1 per feature class) is error free as determined by the USGS MP tool Metadata content contains sufficient detail and all pertinent information regarding source materials,
projections, datums, processing steps, etc. Content should be consistent across all feature classes.
DEM Checklist Overview
Correct number of files is delivered and all files are in IMG format Run “QCRaster” script to ensure all files are tiled 1500m x 1500m with a 1 meter resolution Check sample of rasters to ensure projection/coordinate is accurate with project
specifications (NAD83 – PA11, UTM Zone 4 and Ellipsoid Heights). All units shall be in meters to 2 decimal places.
Review
Manually review bare-earth DEMs with a hillshade to check for issues with hydro-flattening process or any general anomalies that may be present. Specifically, water should be flowing downhill, water features should NOT be floating above surrounding terrain and bridges/box culverts should NOT be present in bare-earth DEM. Hydrologic breaklines should be used during review of DEMs to ensure features are carried through bridges to ensure all hydrographic features are flattened in the DEM. However, breaklines should be flipped off and on for each window display as breaklines turned on will hide floating hydrographic features. The tile grid should usually be turned off as well as it can hide anomalies along borders between tiles.
Dewberry’s DEM QC SOP and standardized DEM edit calls should be used.
Oahu LiDAR QA Report May 16, 2014 Page 31 of 31
Metadata
Project level DEM metadata XML file is error free as determined by the USGS MP tool Metadata content contains sufficient detail and all pertinent information regarding source materials,
projections, datums, processing steps, etc.